Lawrence Berkeley National Laboratory
Lawrence Berkeley National Laboratory
Title
"Photocatalytic generation of hydrogen from water using a cobalt
pentapyridine complex in combination with molecular and
semiconductor nanowire photosensitizers"
Permalink
https://escholarship.org/uc/item/07w48814
Author
Sun, Yujie
Publication Date
2012-08-30
DOI
10.1039/C2SC21163G
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
DISCLAIMER
This document was prepared as an account of work sponsored by the United States
Government. While this document is believed to contain correct information, neither
the United States Government nor any agency thereof, nor The Regents of the
University of California, nor any of their employees, makes any warranty, express or
implied, or assumes any legal responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process disclosed, or
represents that its use would not infringe privately owned rights. Reference herein to
any specific commercial product, process, or service by its trade name, trademark,
manufacturer, or otherwise, does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States Government or any
agency thereof, or The Regents of the University of California. The views and
opinions of authors expressed herein do not necessarily state or reflect those of the
United States Government or any agency thereof or The Regents of the University of
California.
Photocatalytic generation of hydrogen from water using a cobalt
pentapyridine complex in combination with molecular and
semiconductor nanowire photosensitizers
Yujie Sun,
a,c
Jianwei Sun,
a,d
Jeffrey R. Long,
a,d,e
* Peidong Yang
a,e
* and Christopher J. Chang
a,b,c
*
Recently, a family of cobalt pentapyridine complexes of the type [(R-PY5Me
2
)Co(H
2
O)])(CF
3
SO
3
)
2
, (R =
CF
3
, H, or NMe
2
; PY5Me
2
= 2,6-bis(1,1-di(pyridin-2-yl)ethyl)pyridine) were shown to catalyze the
electrochemical generation of hydrogen from neutral aqueous solutions using a mercury electrode. We
now report that the CF
3
derivative of this series, [(CF
3
PY5Me
2
)Co(H
2
O)](CF
3
SO
3
)
2
(1), can also operate 10
in neutral water as an electrocatalyst for hydrogen generation under soluble, diffusion-limited conditions
on a glassy carbon electrode, as well as a photocatalyst for hydrogen production using either molecular or
semiconductor nanowire photosensitizers. Owing to its relatively low overpotential compared to other
members of the PY5 family, complex 1 exhibits multiple redox features on glassy carbon, including a
one-proton, one-electron coupled oxidative wave. Further, rotating disk electrode voltammetry 15
measurements reveal the efficacy of 1 as a competent hydrogen evolution catalyst under soluble,
diffusion-limited conditions. In addition, we establish that 1 can also generate hydrogen from neutral
water under photocatalytic conditions with visible light irradiation (
irr
455 nm), using [Ru(bpy)
3
]
2+
as a
molecular inorganic chromophore and ascorbic acid as a sacrificial donor. Dynamic light scattering
measurements show no evidence for nanoparticle formation for the duration of the photolytic hydrogen 20
evolution experiments. Finally, we demonstrate that 1 is also able to enhance the hydrogen photolysis
yield of GaP nanowires in water, showing that this catalyst is compatible with solid-state photosensitizers.
Taken together, these data establish that the well-defined cobalt pentapyridine complex
[(CF
3
PY5Me
2
)Co(H
2
O)]
2+
is a versatile catalyst for hydrogen production from pure aqueous solutions
using either solar or electrical input, providing a starting point for integrating molecular systems into 25
sustainable energy generation devices.
Introduction
The combination of rising global energy demands, diminishing
fossil fuel stores, and climate change has prompted intense
interest in developing alternative carbon-neutral energy 30
technologies. Harnessing solar energy to synthesize sustainable
chemical fuels is a promising solution to the emerging energy
challenge.
15
An appealing approach to this ultimate goal is to
drive chemical water splitting to hydrogen and oxygen using
solar energy input,
6
since the generation and combustion of 35
hydrogen from water is carbon neutral and sunlight is a
sustainable energy source.
A key challenge for water splitting is developing catalysts for
the direct and efficient production of hydrogen from protons.
Platinum and other heterogeneous precious metal catalysts have 40
been studied for hydrogen generation for decades, but ultimately
suffer from high cost and low abundance.
79
Alternatively, solid-
state catalysts composed of earth-abundant elements, such as
metal alloys and molybdenum sulfides, have also been
investigated.
5,1019
However, it remains a challenge to rationally 45
assess precise structure-activity relationships in these
heterogeneous systems. Enzymes like hydrogenases that utilize
earth-abundant metals offer an attractive approach to hydrogen
evolution catalysis with high catalytic activity and efficiency;
however, it is difficult to integrate hydrogenases into solar water-50
splitting devices owing to their large size and relative long-term
instability. Enzyme mimics provide exquisite insight into
biological systems and lay the groundwork for new catalyst
design principles, but they often function in organic solvents with
organic acids.
2023
As such, a growing number of abiotic 55
molecular catalysts for electrocatalytic and photocatalytic
hydrogen production featuring earth-abundant elements,
24
including iron,
2527
cobalt,
2837
nickel,
3842
and molybdenum,
4345
are being reported.
An important step forward for the field of H
2
catalysis is the 60
development of systems that can operate in aqueous media,
because using water as both the substrate and solvent increases
substrate concentration while minimizing organic additives and
waste by-products. As a molecular approach toward this end, we
recently reported that a cobalt pentapyridine complex, 65
[(PY5Me
2
)Co(H
2
O)](CF
3
SO
3
)
2
(2; PY5Me
2
= 2,6-bis(1,1-
Fig. 1 Molecular cobalt pentapyridine complexes for catalytic H
2
generation.
di(pyridin-2-yl)ethyl)pyridine; see Fig. 1) is a robust and efficient
electrocatalyst for hydrogen evolution in pH 7 buffer on a 5
mercury electrode, albeit at fairly high overpotential.
46
However,
the tunability of the PY5Me
2
platform allowed us to modify the
para-position of its central pyridine and synthesize the derivative
[(CF
3
PY5Me
2
)Co(H
2
O)](CF
3
SO
3
)
2
(1), which showed a positive
shift in both the Co(II)/Co(I) reduction potential and the 10
overpotential for H
2
catalysis. To demonstrate that the
performance of this catalyst is not restricted to electrochemistry
on mercury, which can adsorb molecular species,
28
we now report
that 1 is indeed a versatile system for electro- and photochemical
H
2
generation in water. In particular, we show that it is a 15
competent electrocatalyst under soluble, diffusion-limited
conditions using a glassy carbon electrode, and, importantly, that
it can function as a photocatalyst in combination with either a
simple molecular chromophore such as [Ru(bpy)
3
]
2+
or a
semiconductor GaP nanowire photosensitizer system.
47
20
Results and Discussion
Catalytic hydrogen generation on a glassy carbon electrode
Since the exogenous ligand L at the apical position of the PY5-
cobalt complexes is exchangeable and sensitive to solvent, we
chose to synthesize 1-CH
3
CN for electrochemical studies in 25
acetonitrile, while using the aquo complex 1 for experiments
conducted in aqueous media. Employing these complexes avoids
possible solvent contamination in electrochemical studies.
Similar to our previous report, metalation of CF
3
PY5Me
2
with
Co(CF
3
SO
3
)
2
(MeCN)
2
in acetonitrile at room temperature results 30
in the formation of 1-CH
3
CN, the crystal structure of which is
shown in Fig. S1.
46
In agreement with the reported structure of 2-
CH
3
CN, the Co(II) center in 1-CH
3
CN resides in a slightly
distorted octahedral geometry with acetonitrile bound at the
apical site. The structure of 1 has been reported previously.
46
35
The cyclic voltammogram of 1-CH
3
CN in acetonitrile solution
features two reversible redox processes at E
1/2
= 0.98 V and 0.64
V vs SHE, assigned to metal-based Co(III)/Co(II) and
Co(II)/Co(I) couples, respectively, with another irreversible
reduction peak at –1.57 V vs SHE (Fig. S2). Compared to the 40
redox processes observed for parent PY5Me
2
complex 2-CH
3
CN
in acetonitrile, CF
3
substitution on the para position of the central
pyridine ring positively shifts the formal Co(III)/(II), Co(II)/(I),
and Co(I)/(0) couples of 1-CH
3
CN by ca. 105, 140 and 120 mV,
respectively. As the free ligand CF
3
PY5Me
2
is redox-silent in the 45
same potential region (see Fig. S3), the data suggest that these
observed features are metal-dependent.
Owing to the large overpotential of 2 for hydrogen evolution
catalysis and the relative small electrochemical window of the
glassy carbon electrode in pH 7 aqueous media, no apparent 50
Fig. 2 Cyclic voltammograms of 0.1 mM 1 (solid line) and blank glassy
carbon electrode (dotted line) in 0.1 M phosphate buffer at pH 7. Inset:
pH dependence of the oxidation peak of 1 in 0.1 M buffered electrolytes
at various pH values. Conditions: 0.1 M NaClO
4
added as the supporting 55
electrolyte, scan rate = 100 mV/s, Ar atmosphere.
Fig. 3 Cathodic scans of 0.3 mM 1 in 0.1 M phosphate buffer and 0.1 M
NaClO
4
at pH 7 at different rotating rates: 100 (purple), 400 (blue), 800
(green), 1600 (orange), 2400 (red) rpm, (scan rate: 25 mV/s). Inset: 60
Levich plot of current density at overpotential = 500 mV versus the
square root of rotating rate.
reduction feature of 2 was observed before the rise of the glassy
carbon background current. In contrast, the CF
3
derivative 1
exhibits a well-resolved and irreversible cathodic peak at 0.89 V 65
vs SHE in 0.1 M phosphate buffered to pH 7, with a much more
pronounced rise in current density compared to the background
(Fig. 2). In the positive potential direction, a reversible redox
feature at 0.35 V vs SHE was also observed. The pH-dependent
cyclic voltammograms of 1 in 0.1 M phosphate buffer with 0.1 M 70
NaClO
4
as the supporting electrolyte are shown in Fig. S4. The
oxidation wave shifts positively along the decrease of pH from 9
to 4, with a slope of 58.1 mV/pH (inset of Fig. 2), indicating a
one-proton and one-electron redox process close to the ideal
value of 59 mV/pH. This observation led us to assign the 75
oxidation wave as a Co(II)-OH
2
/Co(III)-OH couple. Similar
results have been reported for other cobalt complexes with
pentadentate ligands in aqueous medium.
35,48
The pH-dependence
of the reduction feature is more complex and is a topic under
current investigation. 80
Fig. 4 Rotating disk electrode voltammograms of 0.1 mM 1 (solid) and
blank glassy carbon electrode (dotted) in 0.1 M phosphate buffer at pH 7.
Inset: n
app
plot of 1 versus potential. Conditions: 0.1 M NaClO
4
added as
the supporting electrolyte, scan rate = 25 mV/s, rotation rate = 400 rpm/s, 5
Ar atmosphere.
Rotating disk electrode voltammetry (RDEV) studies
To demonstrate the molecular nature of cobalt pentapyridine
complex as a hydrogen generation catalyst in aqueous media, a
rotating disk electrode (RDE) was utilized to probe the 10
hydrodynamics of the system in 0.1 M phosphate buffer at pH 7.
Fig. 3 displays the RDE voltammograms of 1 at different rotation
rates with a scan rate of 25 mV/s. A linear Levich plot of the
current density at 0.900 V vs SHE versus the square root of the
rotation rate was obtained (inset of Fig. 3), indicating that the 15
catalytic current is under diffusion control.
In order to compare the performance of 1 to that of other
hydrogen catalysts in aqueous media, we sought to determine its
apparent rate of proton reduction utilizing a method recently
reported by Peters and co-workers.
33
The parameter n
app
, defined 20
as the apparent number of electrons delivered to the catalyst
before it diffuses away from the electrode surface, can be
calculated by the electrocatalytic current density normalized for
the delivery of the catalyst to the surface as illustrated in eq 1.
25
Here, j
p
is the plateau current density for the Co(II)-aqua/Co(III)-
hydroxide couple and j
c
is the catalytic current density as shown
in its RDE voltammogram (Fig. 3). The plot of n
app
versus
potential for 1 is included in the inset of Fig 3. The value of n
app
increases dramatically after the onset of catalysis at ca. 0.8 V vs 30
SHE, consistent with the cyclic voltammogram in Fig. 2, reaching
nearly 16 at 0.900 V vs SHE (overpotential = 487 mV).
Compared to the reported n
app
values of 1-8 for a series of cobalt
complexes at an overpotential of ca. 500 mV in pH 2.2
buffers,
33
which were measured at the same scan rate and rotation 35
rate as this present study, 1 exhibits better efficiency as a
hydrogen evolution catalyst in neutral pH aqueous solution at
approximately the same overpotential. Similar results were
obtained in the 0.2 M NaClO
4
(Figs S5 and S6), indicating that
phosphate does not play a special role in this system. In addition, 40
a Faradaic efficiency of 95 10 % was measured by gas
Fig. 5 Controlled potential electrolysis of 0.1 mM 1 (solid) and rinsed
glassy plate after bulk electrolysis of 1 (dotted) in 0.1 M phosphate buffer
and 0.1 M NaClO
4
at pH 7, showing charge build-up versus time with an 45
applied potential of 0.963 V vs SHE (overpotential = 550 mV).
chromatography for a 3-h bulk electrolysis of 1 at an applied
potential of 0.963 V vs SHE (overpotential= 550 mV) in pH 7
buffer (Fig. 5). The used glassy carbon plate was rinsed with
water and used as the working electrode in fresh buffer for 50
another blank bulk electrolysis, which passed much less charge
under the same condition (Fig. 5). These results demonstrate the
efficacy of 1 as an efficient and robust molecular electrocatalyst
for hydrogen evolution reaction.
Photocatalytic hydrogen production using a molecular 55
photosensitizer
After establishing that the cobalt pentapyridine complex 1 is a
competent molecular electrocatalyst for hydrogen evolution in
neutral water under diffusion-limited conditions, we next tested
whether it could also be utilized as a photocatalyst under similar 60
conditions. To this end, we carried out photocatalysis
experiments in neutral aqueous solutions using [Ru(bpy)
3
]Cl
2
as a
soluble molecular inorganic photosensitizer and ascorbic acid as
an electron donor. As shown in Fig. 6a and Fig. S7, upon
irradiation with light of wavelength
irr
455 nm at room 65
temperature, a solution of 50 M catalyst, 0.2 mM [Ru(bpy)
3
]Cl
2
,
and 0.1 M ascorbic acid in 1.0 M phosphate buffer at pH 7
evolves hydrogen. The hydrogen evolution rate is initially linear
in the first 2 h, followed by a slight deviation, until reaching the
plateau of ca. 0.5 mL after 8 h of photolysis. In a separate 70
experiment, we tested the stability of catalyst 1 and the
chromophore during photolysis. As shown in Fig S8, after 10 h
photolysis, further addition of 0.2 mM chromophore resumes
nearly 40% activity of 1 under another 4 h illumination,
indicating that although some of the catalyst may deactivate 75
during the first 10-h photolysis, chromophore decomposition is
the primary reason for the cease of hydrogen evolution after 8-h
illumination.
35
Importantly, control experiments without
photosensitizer or ascorbic acid showed no H
2
generation, and in
the absence of catalyst, only negligible hydrogen was detected 80
under the same conditions (Fig. 6a and Fig. S7), establishing that
all three components are necessary for the efficient photocatalytic
evolution of hydrogen from water.
35
Compared to the
photocatalytic hydrogen evolution catalyzed by 2 and 3 (Fig. 6a),
![Fig. 6 (a) Photocatalytic hydrogen evolution over time as measured by gas chromatography for aqueous solutions containing 50 M 1 (filled circles), 50 M 2 (filled squares), 50 M 3 (filled triangles), or no catalyst (open circles) with 0.2 mM [Ru(bpy)3]Cl2 and 0.1 M ascorbic acid in 1.0 5 M phosphate buffer of pH 7. (b) Photogenerated hydrogen volume after 2 h of illumination (irr 455 nm) of 0.2 mM [Ru(bpy)3]Cl2 and 0.1 M ascorbic acid in 1.0 M phosphate buffer of pH 7 with various concentrations of 1. (c) Photogenerated hydrogen volume after 2 hour illumination (irr 455 nm) of 50 M 1 and 0.1 M ascorbic acid in 1.0 M 10 phosphate buffer of pH 7 with various concentrations of [Ru(bpy)3]Cl2. The light source was a 150 W Xe lamp coupled to a 455-nm long-pass filter.](/figures/fig-6-a-photocatalytic-hydrogen-evolution-over-time-as-xz2vdcro.webp)
![Fig. 7 Particle size distribution of photolysis solutions containing 50 M 1, 0.2 mM [Ru(bpy)3]Cl2, and 0.1 M ascorbic acid in 1.0 M phosphate buffer of pH 7 after illumination (irr 455 nm) of 0 h (a), 0.5 h (b), and 3 h (c), determined by dynamic light scattering measurements.](/figures/fig-7-particle-size-distribution-of-photolysis-solutions-37fylg81.webp)
